The giant magnetoresistance effect that changes with magnetic field (awarded the Nobel Prize in Physics in 2007) and the resistive switching effect (or memristor) that changes with electric field are both non-volatile changes in the resistance state of a material or structure by applying an external field. They form the physical basis of many semiconductor chips and sensors. Due to the limitations of magnetic storage imposed by the superparamagnetic effect, memories based on switching effects (or memristors) are recognized as the most promising next-generation memories. Due to their unique resistive characteristics, memristors have great potential in the field of artificial intelligence, such as in the development of analog synaptic devices and neuromorphic computers.

Currently, the operation of memristors is mainly based on electrical manipulation, and most of the optical manipulation or control of memristors is volatile. Non-volatile optical control of memristors can only be preliminarily achieved in some photosensitive materials such as phase-change or polarization materials, and the control effect is very limited. Thus, achieving direct voltage modulation (photovoltage, photovoltaic effect) of memristors purely with light, without relying on electrical effects, has been a highly challenging issue in this field. Therefore, solving this problem will provide support for the construction of visual neural networks involved in the field of artificial intelligence, the development of new all-optical controlled sensor chips, new memories, and especially the breakthrough of bottlenecks in this field.


 Prof. Hui Wang

School of Physics and Astronomy, SJTU


        2023.4.26 12:00-13:30